Peter Norvig in his essay "Teach Yourself Programming in Ten Years" says, "If you want, put in four years at a college (or more at a graduate school). This will give you access to some jobs that require credentials, and it will give you a deeper understanding of the field, but if you don't enjoy school, you can (with some dedication) get similar experience on the job. In any case, book learning alone won't be enough. "Computer science education cannot make anybody an expert programmer any more than studying brushes and pigment can make somebody an expert painter" says Eric Raymond, author of The New Hacker's Dictionary. One of the best programmers I ever hired had only a High School degree; he's produced a lot of great software, has his own news group, and made enough in stock options to buy his own nightclub. # Work on projects with other programmers. Be the best programmer on some projects; be the worst on some others. When you're the best, you get to test your abilities to lead a project, and to inspire others with your vision. When you're the worst, you learn what the masters do, and you learn what they don't like to do (because they make you do it for them)."

He is referring to Jamie Zawinski. http://norvig.com/21-days.html

Two points: 1) It does seem to matter for at least CS grad school, as was said in Undergraduation, because of research opportunities. So if you just want to start a web startup and get rich, fine, but if you want to go to grad school after you get rich, not so fine.

2) I've taken classes at CMU, Berkeley, and Stanford, and while the differences didn't particularly correlate with the rankings (since the relative rankings of those three places in CS get permuted regularly anyway), there were serious differences in attitude and rigor. CMU and Berkeley were both very rigorous; the difference was that CMU was rigorous in a somewhat more pointless, soul-sucking manner, on average, and didn't explain things as well. Stanford was somewhat less rigorous, which I imagine was because of the Ivy-league difficult-to-get-in but easy-to-pass phenomenon. This seems to correlate somewhat with the number of undergrads at each place involved in research; at both CMU and Berkeley, I know many undergrads involved in research, while the one professor I talk to at Stanford hungrily scrounges for the rare undergrad who's good at research.

So there are differences in curriculum and selection, I think; they're just not particularly amazing predictors of success. They certainly should influence the decision of which colleges to apply to, though.

Though as a reader I start sympathetic to both the thesis (it matches my priors) and the author (always enjoy PG's essays), this struck me as hand-waving.

For all the talk of, "I have a lot of data" and "[w]e're just finally able to measure it," there are no supporting numbers, just general impressions.

C'mon, give this topic the rigorous Bayesian treatment. How does the population of YC market-successes compare with YC-chosen and YC-applicants, along the 'prestigious college' dimension? How do these compare with the college-graduate population as a whole, and the college-graduate population succeeding in other competitive fields?

Those numbers would be great, and I'm rooting for the "college-doesn't-matter" result.

First, I'd just like to point out that PG is in an interesting place to report that college doesn't matter, having himself attended two Ivy League schools.

There are two arguments happening here. The first is that entrepreneurship and business success are not so well correlated with where you go to undergrad. The second is that where you go to college doesn't make much of a difference in life in general. I won't really quarrel with the first, since it's been demonstrated pretty often, and YC's results support it as well. The second, on the other hand...

The main problem with trying to assert that college doesn't make a difference in life is that there's really no easy measurement you can do over "net change" in life. That's due both to the long span of time a life consumes as well as the complex and/or fuzzy cause-effect relationships that are at play throughout. At best, I think, you would need to argue over objective differences that have some immediate effect, and then hope that those actually do make a difference. As an example, here are some of the things that I think made Paul Graham successful that are pretty strongly related to his going to graduate school at Harvard:

- he met RTM and Trevor there

- he got to spend his time in Boston, where he's acknowledged startups can thrive

- being able to drop the H-bomb almost certainly helped give credibility to his company

- he learned Lisp (not really the most popular language at the time, except at certain places)

Would Viaweb have happened had he gone elsewhere, or skipped grad school altogether? Well, maybe. But I doubt it. Harvard's name is what convened the talent (peers, advisors, mentors, investors) that shaped much of that company, and it's there that Harvard's value lies. I think this essay horribly undervalues the magic that happens when everyone in the world thinks your university is amazing.

When I was taking an education theory class we had to do simulated admissions. We had to do three applicants every ten minutes, which works out to just over three minutes per app. In that amount of time you can quickly scan over the grades and SATs, and maybe read the first paragraph of each essay if you're lucky. The other thing I learned is that the real admissions officers got statistics updates twice a day for the average GPA and SAT score, and also the projected US News rank. This means that whether or not your grades and SATs were good enough depended largely on whether your app was read before or after lunch, because what it took to get into the college completely changed every time they were handed the new report.

For low income minority students there was an option to set the app aside for a second reading in order to learn more about the student's situation and if there were an ameliorating factors, but for the rest of the students the admissions officers were expected to make a decision on the first pass after the three minutes.

Athletes also largely got pre-approved by the academic department they were applying for, so they pretty much knew whether they'd be accepted before they ever applied. The flip side is that they only got an edge in admissions if they applied early decision, because if they were going to bring down the average GPA then they had to bring up the average matriculation in order to not affect the overall US News rank.

I think it's one of those things like eVoting. That is, people with no CS experience think eVoting is totally secure whereas CS experts know it isn't. Similarly, I highly suspect that anyone who thinks getting admitted to an Ivy shows a certain baseline level of respectability has never worked in admissions. I'd guarantee it.

As for the importance of college GPA, if you want to see something funny then apply for a wall street job. If they ask you what your GPA was in college, ask them how GPA correlates with alpha. :-)

The craziest thing was that Google did a massive HR survey and determined that there was basically zero correlation between college GPA and value created for the company. Because of this they decided that they would give jobs to five or six people with sub 3.0 GPAs each year. Well if there is little or no correlation, why should it matter what GPA is at all? I suspect the psychology behind the Google hiring process has a lot in common with the psychology of female circumcision. That is, it was done to me so it must be a good thing. And if it's a good thing, then by definition it must be good to do unto others.

edit: fixed a few grammatical errors

When I was in college, I figured out that the education you get doesn't really change. However, what does change is the people that you are associating with.

At the "elite" colleges, there is a greater percentage of smart students, which lends itself to more productive discussions and more difficult tests that force you to challenge yourself. Also, it is like the difference between the regular classes in high school and advanced classes -- in the advanced classes you cover the same material, but with greater breadth and a deeper understanding.

And the number 1 biggest advantage of going to an elite college is that you have a much better chance of meeting the world's future movers and shakers -- the people that will be the educational and business leaders of the next generation. That alone is a good reason to go to an elite college.

So I have to disagree, and say that there are indeed advantages to going to an elite college, just not the advantages that most people think.

But it DOES matter where you go to college, and the essay itself argues for that.

In this context, we can say there are three things that can determine someone's success:

1) confidence others have in you

2) self-confidence

3) actual ability

Going to a "good" college affects 1 and 2 the most, and arguably 3 the least. In a startup 3 and 2 are arguably the most important, and 1 the least - so I guess that agrees with pg's argument. In the corporate world, 1 and 2 matter the most, and 3 the least - aligned with what a brand-name college provides.

So the argument seems a bit circular: based on the yc sample set, college doesn't matter because college doesn't matter to the yc sample set. Even if that SHOULD extend to the world at large, it doesn't.

I did my D.Phil. at Oxford University, and my B.Sc. at Simon Fraser University; while SFU is one of Canada's best universities, it clearly doesn't have the stature of Oxford or Harvard. Were the students at Oxford more intelligent than the students at SFU? No. Were they more motivated? No. Are there jobs for which I'd rather hire an Oxford alumnus than an SFU alumnus? Absolutely.

The largest advantage of a great university like Oxford is in increasing the range and depth of ideas presented to students. It's impossible to completely separate the teaching role of a university from its research role: Teaching (and inconvenient questions asked by irritating students, like yours truly, in class) informs research, and research informs teaching. And quite apart from the "indigenous" research, if you're at a minor university without a strong reputation for research, you don't have Knuth dropping by for a few months, or Rivest visiting to give a lecture (albeit about the rather odd notion of obtaining forward security by using a stream of broadcasted random numbers which is too fast to be stored).

There are undeniable advantages to being at a great university in being exposed to new ideas. Why haven't PG et al. noticed this? Probably because their measurement -- can someone create a successful startup? -- is just as biased as everyone else's measurements. After all, when was the last time that a YC-funded startup really did anything new?

I have thought about this a lot.In fact lot of my teen years were spent thinking about this. I went to an elite school in India. And I leave a comment here just to confirm that its the same all over the world. These lines are gold. ( I mean very valuable :)

"The unfortunate thing is not just that people are judged by such a superficial test, but that so many judge themselves by it. A lot of people, probably the majority of people in the America, have some amount of insecurity about where, or whether, they went to college. The tragedy of the situation is that by far the greatest liability of not having gone to the college you'd have liked is your own feeling that you're thereby lacking something. Colleges are a bit like exclusive clubs in this respect. There is only one real advantage to being a member of most exclusive clubs: you know you wouldn't be missing much if you weren't. When you're excluded, you can only imagine the advantages of being an insider. But invariably they're larger in your imagination than in real life."

Analogue: no one ever gets penalized for upvoting comments that have already been upvoted several times.

Same basic idea.

See also: http://weblog.raganwald.com/2005/07/why-you-need-degree-to-w...

PG: Here's a good way to test this: Please ask the next round of applicants to NOT mention the school they went to.

What do you think?

How was the data analysed?

The most obvious way, and the way the essay seems to imply, is simply to look for correlations between success / failure and which school the founders attended.

If this is how it was done, then the conclusion is flawed. All it shows is that the school doesn't matter, conditional on being accepted by YC (rather than rejected). This is a completely different conclusion than the one the essay reaches, namely that the school doesn't matter, period.

As always, a convincing case, well edited. :) Paul's point probably plays even better among Hacker News readers, who are probably more suspicious than the average bear of college in general.

Seems to me there's an extra factor contributing to the longevity of the "which school" criterion. There are other roles in the business (especially larger businesses) such as biz dev, M&A, etc. where attending an elite school might enhance success considerably - not for the academic content learned, but for the network of friends and colleagues you gain at that school. It's the English Grammar School model of corporate success.

s Paul points out, hiring decisions are often made by HR types: but HR is a very extroverted, business process focussed profession. As personalities, and in job description, the HR folks have more in common with the people and roles that do benefit from elite school connections than with technical people for whom the elite school is irrelevant. Like all of us, they apply lessons from familiar situations to more unfamiliar situations. And so the Stanford Screen lives on.

"The curve for faculty is a lot flatter than for students, especially in math and the hard sciences; you have to go pretty far down the list of colleges before you stop finding smart professors in the math department."

I wish this would be repeated more often. I've implemented many interesting algorithms based on papers written by people at, like, Louisiana State University. Noname places.

This essay seems to resonate with another Paul Graham theme -- to avoid prestige while choosing a job. His point then is that you are never compensated for prestige, so you might as well stay away from it.

It looks to me like the education market is starting to catch up with the job market. Prestige in either area probably means that you're paying too much or being paid too little.

I'm surprised it took Paul this long to realize this about universities. A little less than half way through high school I realized that the entire process (get good grades to go to a good school, get good grades there so you can get a good job) is crap. I decided that it was more important to spend my time in high school learning what I cared about then to obsess over GPA.

I never got around to finishing my college degree, I left about 2 years into it and got a job because I ran out of things to learn in college. I've worked for a series of both small and large companies and I've found that most companies also don't care where one went to college (Google is a big exception to this). Most companies simply don't value people that are fresh out of college because colleges don't teach people the right skills.

It's true however that in good schools there are more smart people and you are more likely to meet interesting people, but that doesn't mean you need to go to a top school, just live close to one.

I was amazed initially (though that quickly faded) how few truly "smart" undergraduates there were at MIT in the 9 years I hung around there. I put smart in quotes because I really don't like that word, but I don't have a great substitute either. Obviously they were smart enough to get in. But most were significantly lacking in at least one necessary area to be labeled "smart" IMO (often several), including intuitiveness, creativity, being a self-starter, analytical thinking, breadth and depth of knowledge, intellectual curiosity, etc.

Putting exact numbers on it is difficult due to the self-segregation at MIT and subjectivity obviously at play. However, I would say perhaps 30% fall into the seriously lacking category, another 30% in a substantially lacking category and another 25% in a good but not great category (perhaps lacking only one major thing). That leaves only 15% or so of people I came across I would say were "smart" in the sense described above. When you move to people with the essential qualities necessary to work successfully at a really early stage startup, I would say it drops to 7% or so. I'm not saying found one, just work at one as one of the first employees. When you move to people with all the founding qualities, the % probably drops to 1-2%. I know this first hand after trying to start companies with people in the "smart" category.

The above has really hit home other times as well. Here are two more occasions. First, when graduating, 62 people in my class out of about 1,080 (~6%) were inducted into Phi Beta Kappa. I was shocked at a) how many of these people were my friends at MIT; b) how many of these people I had come across (given that I didn't meet all 1,080; and c) that this list included pretty much all of the people I had previously labeled "smart" myself in my class that I had come across.

Second, my wife taught statistics at MIT for several years. One time a kid came up to her complaining about an exam grade and told her it was not possible to have gotten an A on this exam. He said all his friends studied real hard and yet they all ended up with Cs and Ds. To which she replied, actually roughly 30% of the class got As. The point is, they were completely self-segregated by "smartness" and didn't even really know it.

All that being said, an MIT degree really does go a long way in terms of impressing the average person/company. I have done independent consulting and applied for jobs where they basically stopped researching me when they saw MIT. That is the rule, not the exception. And that was in Boston. Now that I live in PA, it seems even more shocking to people to come across someone from MIT. Given the above, and in agreement with PG, this behavior is unwarranted. I of course am thankful, but it also annoys me that so many people are free-riding on the name.

For me the biggest thing about going to an elite school is expectation.

The more you expect for yourself, the more others expect from you the more you are driven to succeed or make something of your life.

Elite colleges are like the newsweek article on octopart, it makes you want to beat the expectations on you, only several order or magnitudes less.

I think getting an education is secondary in college. Finding people who live with passion used to be a problem when we didn't have the internet. And working/studying in elite institutions gave you a higher chance of meeting passionate people.

I've noticed the same thing about top-tier chemists. Grad schools tend to have a good mix of Ivy kids and others, but the most brilliant scientists always seem to come from some mid-level state university or a little liberal arts college.

Success at the highest levels of scientific research takes an odd mix of creativity, intuition, and intellegence that has little to do with academic success in high school or undergrad.

I think Paul's logic here is subtly backwards...

"It doesn't matter much where a given individual went to college."

From Y-Combinator's perspective, yes, I'm sure that's true. The people they're looking at are self-selected, and pretty much guaranteed to be smart, motivated and well-educated, which means that all Y-Combinator needs to do is worry about how they think - not where they were taught. But that's after the fact. To someone deciding what college to go to, or where to send their children, I think the choice is still very important.

"someone with a real thirst for knowledge will be able to find a few smart people to learn from at a school that isn't prestigious at all." True, of course. But where does that thirst for knowledge come from? With the right support, or the right teachers, perhaps that party animal would have changed his ways. As far as I'm concerned, _that_ is really the purpose of good schools.

In Y-Combinator terms, it's not that they make a founder more likely to succeed: it's that they make him more likely to try being a founder in the first place.

One hundred years from now will an "elite college" still exist? The reputations of today's elite colleges stem from a perceived legacy of alum achievement 25, 50, 100 years ago. Are factors operating now that will more evenly distribute this achievement in the future? Or is this "legacy of achievement" and conferred eliteness already BS?

I liked this article, if for the simple fact that it can only help my cause (I graduated, though not from a prestigious school)!

The whole time I was going to school I was super impatient to get done so I could actually get working on something real. I was really just after the piece of paper they give you at the end (though I never attended my graduation to get that paper, and the original was sent to the wrong house).

Once I graduated I started my own project so that I would have some work to show when applying for jobs. I found that that experience led to as much or more knowledge gained as my best year in school. And that is not a knock against the school I went to (I thought the staff there was awesome).

I am always reminded of a quote from some movie (cant think of the name of it now)... "You spent x thousands of dollars on an education you could have got for 2 dollars in library late fees"

The second paragraph reminded me of this Alan Watts talk: http://www.freshminds.com/animation/alan_watts_life.html.

But unlike the Watts talk, the essay does converge to a very rigid notion of "success".

"Good" colleges are a self-fulfilling prophecy. If the best students organized online and randomly picked some backwoods institution and applied there, it would instantly become the place to be. Meanwhile the previous top schools would be left with the second-stringers.

It bears repeating: out of the top 10 richest "U.S. Americans", 4 inherited it; of the six who didn't, one has a degree (Warren Buffett), and 5 are drop-outs: Bill Gates, Sheldon Adelson, Larry Ellison, Paul Allen, and Michael Dell. Woz dropped out to start Apple (he went back to school only after he was done working there) and Jobs dropped out of Reed College here in Portland. (We're all very proud of him.)

Anyone have that awesome quote from Bowling for Columbine where Moore is interviewing South Park co-creator Matt Stone and Stone explains the supposed consequences of not getting into honors classes in the 3rd grade?

It doesn't take a Paul Graham to realize this nor do I think this is the first time the thought hit him. There are also things that I, and others, would argue about, in favor or not, but nonetheless,

thank you for writing this.

As someone who took 1 term of a bad C++ class at a community college, and primarily studied Italian prior to dropping out of the University, I can't help but vote this one up.

I always wondered if boso getting into YC had anything to do with the founders having gone to Oxford.

PG: have you noticed that a) GPA matters (at any school); b) that GPA matters at so-called elite schools in particular; c) any other consistencies with regards to college activities and/or honors, e.g. Phi Beta Kappa, having had started clubs or groups, etc.?

the idea that college is necessary is a relatively recent aberration that from what I can tell dates back to the end of WWII and the GI Bill. before that, going to college was rare.

college is not for everyone (we need good plumbers and auto mechanics, fer chrissakes). i didn't apply myself as well as i could have at college, but about 40% of the value I got was from reading interesting books I found in the library, which I could have got for $100 a year while working in the real world and gaining money and experience -- which is what I'm doing now.

that said, some things are very hard to learn outside of a classroom and the mathematical maturity I gained from taking math classes will serve me for a life time.

Here's another dirty little secret: Your grades don't matter either.

I still chuckle when I think of the drones in my fraternity that missed the best 4 years of their lives because their heads were stuck in their books. Sure, they got 4.0 averages, but who cares now?

This is the first of PG's essays I've seen with comments at the bottom. It seems they're brought to us courtesy of Disqus, a YC startup.

http://paulgraham.disqus.com/news_from_the_front/

ha, this is true. i go to rutgers for 4k/yr after a certain ithaca ivy wanted 43k/yr or whatever it was, so i get to keep a college fund's worth of seed money for whatever i do post-grad. and the "elite college" types i meet in the summers are no better-educated than me (and incidentally, drink a lot more...)

undefined

From what I can tell, being a YC alum has similar benefits to being a Harvard (or other Ivy League) alum.

Thank you, this helps me a lot.

undefined